@charon-fan

commit-helper guides users through conventional commits formatting with type/scope/subject structure, includes scope taxonomy by domain, and provides good/bad examples for reference.

figma-designer extracts design systems from figma files and generates detailed visual specifications, tokens, and component hierarchies for implementation. covers workflow from url input through responsive behavior and interaction states.

long-task-coordinator manages multi-session work with durable state and recovery semantics, targeting tasks that span turns, involve handoffs, or require explicit waiting rather than polling.

refactoring-specialist catalogs 10 common code smells with before/after examples and basic refactoring principles, but lacks concrete procedural depth for applying techniques to unfamiliar codebases.

performance-engineer covers systematic diagnosis and optimization across database, api, frontend, and network layers with concrete examples and measurable targets, but lacks explicit failure recovery and assumes ideal conditions.

create-pr automates pull request submission with bilingual documentation sync. guides users through change analysis, documentation updates for english and chinese readmes, and pr creation via conventional commits and gh cli.

planning-with-files provides a persistent markdown-based workflow for multi-step tasks, using a three-file pattern (task_plan.md, notes.md, deliverable.md) to maintain state across context resets and reduce goal drift in agent-driven work.

prd-planner creates product requirement documents using file-based persistence to maintain coherence across context switches. trigger coverage is strong but the actual planning procedure remains underspecified.

code-reviewer provides a templated framework for analyzing pull requests across correctness, security, performance, and testing dimensions. executes via git commands and outputs structured feedback with severity tiers, but lacks explicit decision logic and failure mode handling.

deployment-engineer covers ci/cd pipeline setup, deployment strategies (blue-green, rolling, canary), and environment configuration. lacks actionable procedures, edge case handling, and recovery guidance.

security-auditor provides grep-based pattern matching across owasp top 10 categories with a checklist structure. lacks actionable remediation steps, detailed decision logic, and realistic failure mode handling.

architecting-solutions guides requirement analysis and PRD generation through structured workflows including codebase analysis, constraint identification, and multi-option design comparison. primarily targets software feature design and refactoring planning with emphasis on avoiding over-engineering.

documentation-engineer provides templates and checklists for readme, api, and code documentation. lacks actionable procedures and doesn't address failure modes like outdated docs or missing schema validation.

session-logger captures conversation history to markdown files with flexible triggering in chinese and english. provides a template structure but lacks concrete implementation details and error handling.

api-documenter generates openapi/swagger specifications but lacks actionable procedures, edge case handling, and actual documentation workflows. covers trigger conditions and basic structure only.

api-designer covers rest and graphql architecture with resource-oriented foundations, but lacks concrete procedural depth and failure scenario coverage needed for production api decisions.

qa-expert outlines risk-based testing frameworks and quality gate setup. coverage is shallow, lacks concrete implementation steps for test automation or metrics definition.

auto-trigger is a configuration schema for workflow hooks between skills. it defines event-driven coupling but lacks actionable procedures, error handling, and concrete implementation guidance.

A universal self-improving agent that learns from ALL skill experiences. Uses multi-memory architecture (semantic + episodic + working) to continuously evolve…

Intelligently routes user requests to the most appropriate Claude Code skill. ALWAYS use this skill FIRST when user asks for help, mentions "skill", "which",…

Test automation framework expert for creating and maintaining automated tests. Use when user asks to write tests, automate testing, or improve test coverage.

Implement PRDs/specs with a mandatory precheck review before coding. Use when a user asks to implement a PRD/feature spec/requirements doc or says "implement…

Automatically coordinates multi-skill workflows and triggers follow-up actions. Use when completing PRD creation, implementation, or any milestone that should…

debugger — an installable skill for AI agents, published by charon-fan/agent-playbook.